Antispam that learns from every user
SpamFoo checks each message before it reaches the inbox, on your own server or on an isolated instance we host for you. It catches spam by what the message asks for, and each user's corrections take effect right away.
01Intent
Catch spam by what it asks for, not the words it uses
Keyword rules and word-count filters look for wording they have seen before. Spammers know this and change the wording until a message gets through. SpamFoo's AI engine looks at the request behind the words, along with the links, the sender and the authentication results.
- A fee to release a parcel is still a fee after the subject changes
- The same request is recognized when it's translated into another language
- Links are compared with where they really lead
- SPF, DKIM and DMARC results are part of every decision
02Learning
One click changes how the next message is handled
When a user marks a message as spam or moves one out of junk, SpamFoo applies that choice to their mailbox right away. Similar mail from that sender follows the user's choice from then on. The correction also helps improve SpamFoo for every other user.
Users can see what SpamFoo learned from them on their My Rules page, and remove anything they want to undo. How one correction teaches a mailbox
03Reputation
Sender reputation from the whole SpamFoo network
The SpamFoo network keeps track of how each sending domain and IP address behaves. When a new source of spam shows up at one installation, the rest know about it soon after. Each SpamFoo installation keeps its own copy. When SpamFoo runs on your network, reputation checks keep working if the internet connection drops.
See how many senders are tracked04Your rules
Your allow and block rules always come first
Admins set rules for the whole server or for one domain. Users set their own. Rules can match one sender, a whole domain, a wildcard pattern or an IP range, and SpamFoo follows them before anything else.
05Where it runs
Run it on your own server, or let us host it
SpamFoo Client runs on standard CPUs with no GPU required, on the same machine as your mail server or on another server you control. If your mail server can't spare the CPU and memory, we'll run an isolated SpamFoo instance for you instead. Either way, verdicts come from SpamFoo's own AI engine.
- Windows Server and the major Linux distributions
- Docker images for Linux hosts
- Signed automatic updates
- No MX changes
- Hosted plans with nothing to install
What reading for intent catches
These attacks change their wording all the time, but each one still has to ask the reader for something.
Reworded phishing
A request for a password is still a request for a password, however the subject line is written.
Business email compromise
A short note from the "CEO" asking for a wire transfer or gift cards. How BEC works.
Social engineering
Messages that build trust or urgency so the reader acts before thinking it through.
Financial scams
Fake invoices, overdue fees, refunds that need "confirming" and prizes that need a payment.
Impersonation
Mail that borrows the name of a person or company the reader already knows.
New campaigns
A scam nobody has reported yet can still be judged by what it asks the reader to do.
Anti-Phishing is included
The most expensive email attacks often have no bad link and no attachment. They borrow a name your staff trust. Every SpamFoo license includes Anti-Phishing, which catches them in the same service, with the same dashboard.